Long-life Asphalt Pavements_technical2007

8/11/2019 Long-life Asphalt Pavements_technical2007

1/24

Long-Life Asphalt Pavements Technical version June 2007

1

Long-Life Asphalt Pavements

Technical version


2/24


2

European Asphalt Pavement AssociationRue du Commerce 771040 Brussels,Belgium

[email protected] 2007


3/24


3

Content page

1. Introduction 41.1. Background 4

1.2. Definition of long-life pavements 5

2. Construction of new pavements 62.1. The pavement structure 6

2.1.1. Surface course 72.1.2. Binder course 82.1.3. Base course 82.1.4. Unbound materials and foundation 8

2.2. Design 82.2.1. Fatigue cracking 102.2.2. Permanent deformation 112.2.3. Other mechanisms 112.2.4. Improved design methods 11

2.3. Construction 12

2.4. Maintenance 13

3. Economic aspects 13

4. Future developments 144.1. Procurement and specifications 144.2. Improved economic models 14

5. Summary and conclusions 15

6. Literature 16

Appendix: Durability of surface layers in Europe 18A.1. Introduction 18A.2. Asphalt mixtures 18A.3. Basic assumptions for determining the durability of the surface layers 20

A.4. Results 21


4/24


4

1. Introduction

An important goal in the pavement industry is to provide constructions at ever lowerlife-cycle costs, a goal requiring continuous development in areas such as design,material characterization and production, as well as maintenance techniques and

management. During the last couple of years, the concept of Long-life pavements1

(LLP) has been established in Europe and America. The main purpose of this conceptis to develop a framework, in which more cost-effective pavements are produced. TheLLP-concept is to some extent based on earlier concepts, such as full-depth, stage-by-stage and deep-strength designs, but mainly on more recent advances in materialtechnology, design and functionality. In practice, the intention of the LLP-concept isto significantly extend current pavement design life by restricting distress, such ascracking and rutting, to the pavement surface. Common distress mechanisms such asbottom-up fatigue cracking, rutting in the unbound layers and frost heave should, inprinciple, be completely eliminated.When the availability of road lanes and the road user delay costs are taken into accountin the cost-benefit analyses of pavements it will show that a number of highways indensely populated areas require low maintenance pavements. That means a totalreconstruction of a road pavement is hardly possible or even impossible.This is one of the reasons for developing the Long-Life Pavement concept.

The purpose of this paper is to provide a state-of-the-art review of developments inlong-life asphalt pavements.

1.1 Background

All elements of long-life pavements are not by any means new and some have beencirculating in different forms for a considerable time. For example, several countrieshave utilised so-called full-depth asphalt pavements consisting of relatively thickbitumen-bound layers directly constructed on the sub-grade or placed on relativelythin granular base courses. One of the main arguments for this type of design is thatthe overall pavement thickness can be kept relatively thin compared to conventionaldesigns consisting of relatively thin bitumen-bound layers on relatively thick granularlayers. In addition to this obvious advantage, during the years it has been recognizedthat full-depth and deep-strength pavements often have shown design livessignificantly longer than initially expected [1.]. Similar experiences to those

mentioned have been recognised in the UK, where investigations on motorwayssuggest that fatigue cracking and structural deformation are not prevalent in pavementswhich are constructed with sufficient strength (above a certain level). The mainconclusions suggest that well-constructed flexible pavements show very long servicelives provided that distress mechanisms such as rutting and cracking are confined tothe pavement surface, which significantly facilitates maintenance [2. and 3.].

During 1998/1999, the Conference of European Directors of Roads (CEDR) declared,as suggested by the UK Highways Agency, that so-called Long-Life Pavements werean appropriate area for co-operative research. Subsequently, a working group called

1 It should be emphasised that the term long-life pavements given in this paper is a collective name incorporatingmany of the features shared by concepts such as perpetual, long-lasting, heavy duty and maintenance freepavements.


5/24


5

ELLPAG (European Long-Life Pavement Group) was established as a focal group fordetermining the way forward. The objectives of ELLPAG include producing state-of-the-art reports on European knowledge on the design and maintenance of long-lifepavements. In the long run, a user-friendly comprehensive Best Practice guidancedocument for all types of pavements is intended to be produced [4.].

In the USA, the concept termed Perpetual pavements has rapidly gained acceptancefollowing workshops on the concept held in the first years of the new millennium. Oneof the publications arising from these is a synthesis published by the AsphaltPavement Alliance (APA), a coalition of the National Asphalt Pavement Association(NAPA), the Asphalt Institute (AI) and the State Asphalt Pavement Associations(SAPA) [5.]. Relevant material presented at the Transportation Research Board (TRB)Annual Meeting in 2001 was published in Transportation Research Circular 503Perpetual Bituminous Pavements [6.]. This long-life approach aims at extending the20-year life expectancies of hot-mix asphalt pavements to greater than 50 years [6.].The term Perpetual Pavements is to some extent a rewriting of long-life pavements,but focuses on resisting specific distresses in the each bound layer. In this case, thesurface course, binder course and base courses are selected, designed and performancetested in accordance with deterioration mechanisms such as permanent deformationand fatigue cracking. In addition to long-life and perpetual pavements, several otherapproaches have been used, each comprising different parameters and techniques.

As indicated in this section, the long-life concept is not a single coordinated drivetowards improved pavements. The common feature among different approaches withinthe long-life pavement concept described here is the objective to gather knowledgeregarding design, material characterization and production, as well as maintenancetechniques and management methods to produce pavements with long life in order toachieve low life-cycle costs. Consequently, even though the long-life concept mayappear homogenous, many significant differences arise between differentorganisations and researchers regarding design, material selection and productionmethods. In addition to the two approaches mentioned (long-life pavements andPerpetual Pavements, respectively), some countries, including France, Germany, USAand the UK, already permit design periods in excess of the frequently used 20 years[4.]. However, it should be emphasised that in Europe, solely the UK states that theirpavements are designed for 40 years [4.]. In countries such as the Netherlands, flexiblepavements of motorways are designed in such a way that structural damage will beavoided. In the past they used the Stage-by-stage concept; the existing pavement

was overlaid to become a long-life-pavement. Now the structural design is based onlong life taking into account expected functional use of the road. In the Netherlandsthe traffic intensities are very high and therefore the disruption to traffic has to beavoided. The availability of the road is of utmost importance. The repair of structuraldamage is too costly and too time consuming. In other countries, e.g. France,pavement design life is guided by an economic strategy, which usually is carried outover 30 years. Since most countries possess national pavement design methods, basedon local materials and local experience, only a general strategy for long-life pavementdesign and construction will be described in this document.

1.2 Definition

In the literature several definitions of long-life pavements have been given. In theUSA, perpetual pavements has been closely connected to thick asphalt pavementscomprising a three-layered asphalt pavement including a wear-resistant and renewable


6/24


6

top layer, a rut-resistant and durable intermediate layer and a fatigue-resistant anddurable base layer [6.]. This concept represents an operational definition, or exampleof a long-life pavement, but it restricts long-life pavements to a certain design. Thesynthesis presented by the Asphalt Pavement Alliance (APA) defined a perpetualpavement as an asphalt pavement designed and build to last longer than 50 years [5.].

ELLPAG [4.] uses a similar functional definition:

A long-life pavement is a type of pavement where no significantdeterioration will develop in the foundations or the road base layers

provided that correct surface maintenance is carried out.

The ELLPAG definition above came from the intention to avoid limiting the conceptto a specific life (e.g. 30, 40 or 50 years), which has done in some studies (e.g. [7., 3.and 1.]). In this paper, the functional definition of ELLPAG has been adopted, whichconsequently implies that all pavement layers, except the surfacing, are considered aspermanent. To maintain the overall objective, it is necessary to successively monitorpavement performance in order to ensure that all distress mechanisms are limited tothe surface layer.

Figure 1 illustrates how the goals of the LLP-concept can be approached. The figureillustrates important quantitative components which have been identified by researchersto achieve long-life performance. In addition to the guiding components, somerequisites are important to consider when designing a LLP. Pavement network fundingnormally comes from the government by budget allocations and effective networkmanagement requires that the budget levels are at least sufficient to keep the core roadassets in stable condition in the long term. In the case of LLP, the expected increasedconstruction costs associated with improved materials and construction constitutefinancial restraints even though lower life-cycle costs are obtained in time (see Section3). In addition, the final design of a LLP depends on local conditions such as sub-gradetype, present and future traffic volumes and materials available. From these so-calledprerequisites, the ultimate goal of low life-cycle costs can be obtained based onoptimizing maintenance activities, delay and environmental costs.

Figure 1. Prerequisites, components and goals of long-life pavements concepts

Prerequisites Sufficient financing Knowledge regarding

local conditions- Subgrade types- Traffic volumes- Available materials

Components Design

- Design life, e.g.>40 years- Conservative design criteria

Performance specifications- Wear-resistant top layer- Rut-resistant layers- Fatigue-resistent base course

Periodical resurfacings

Goals Low life-cycle costs

- Few maintenance activities- Low delay costs- Low environmental costs


7/24


7

2. Construction of new pavements

Today, most countries regard the public road transportation system as an essential pre-condition for economic development and welfare. It is therefore important that theasset is managed professionally through all stages including design, construction and

maintenance.

2.1 The pavement structure A flexible pavement consists of the road structure above the formation level, whichnormally comprises bitumen-bound and unbound materials. The pavement structureshould be able to resist the traffic and environment to which it will be exposed in sucha way that structural distress mechanisms are avoided. Normally, pavements aredesigned as layered structures with relatively low strength in the lower layers andmaterials with progressively increasing strengths towards the top (see Figure 2). Therationales for this arrangement are both technical and economic.

Figure 2. Schematic illustration of flexible pavement structure.

In general, the asphalt layers are laid on a bound or unbound road base layer. Startingat the road surface, the first layer is called the surface course. The second layer is thebinder course and the lower layers are the base courses.For a good bearing capacity of the whole road structure it is important that there is agood bond and interlock between all the (bituminous) bound pavement layers and agood interlock between the unbound pavement layers. It of course also applies to theinterlock between the first asphalt layer and the unbound base beneath it.

For the durability of the asphalt structure, it is also essential to have a good bondbetween the asphalt layers that can avoid water penetrating between asphalt layers.

2.1.1 Surface courseThe surface course constitutes the top layer of the pavement and should be able towithstand high traffic- and environment-induced stresses without exhibitingunsatisfactory cracking and rutting in order to provide an even profile for the comfortof the user and at the same time possessing a texture to ensure adequate skidresistance. Depending on local conditions, functional characteristics such as skidresistance, noise reduction and durability are often required for surface courses. In

some cases, rapid drainage of surface water is desired through a porous surface whilein other cases the surface course should be impermeable in order to keep water out ofthe pavement structure. As indicated, the surface layer is important for the pavement

TERRASSYTAUNDERGRUND

BRLAGER

BELGGNING

FRSTRKNINGSLAGER

TERRASSYTATERRASSYTAUNDERGRUNDUNDERGRUND

BRLAGER

BELGGNING

FRSTRKNINGSLAGER

BRLAGERBRLAGER

BELGGNINGBELGGNING

FRSTRKNINGSLAGERFRSTRKNINGSLAGER

Asphalt layers

Unbound road-base

Formation

Subgrade

Unbound sub-base


8/24


8

performance but no single material can provide all the desired characteristics (e.g.porous and impermeable at the same time). A wide range of surface layer products cantherefore be considered appropriate depending on specific requirements:

Asphalt Concrete (AC) Bton Bitumineux Mince (Thin Layers Asphalt Concrete AC-TL) Asphalt Concrete Very Thin Layers (AC-VTL) Ultra Thin Layer Asphalt Concrete (UTLAC) Stone Mastic Asphalt (SMA) Hot Rolled Asphalt (HRA) Porous Asphalt (PA) Double layered Porous Asphalt (2L PA) Mastic Asphalt (MA) Soft Asphalt (SA)

In the end, the selection of the surface course is a matter of identifying the mostappropriate material during the design process. The functional requirements canconflict. For example, noise reduction could require the use of a double-layeredporous asphalt and that conflicts with the requirement of a very durable surface layer.The durability of surface layers can be improved by using higher quality materials.The higher costs of these will be compensated by the lower costs of traffic measuresand user costs.In Appendix A the durability of these surface layers is shown.

2.1.2 Binder courseUnder the assumption that a particular layer solely is subjected to a mode ofdeterioration, long-life pavements are usually designed by preventing permanentdeformation in the bound layers using an appropriate binder course. Binder courses arefrequently adopted on the assumption that the highest shear stresses will occur about50 70 mm below the asphalt surface. The binder course is therefore placed betweensurface course and base courses to reduce rutting by combining qualities of stabilityand durability. Stability can be achieved by sufficient stone-on-stone contact,combined with stiff and/or modified binders.

2.1.3 Base courseThe base course is perhaps the most important structural layer of the pavement, whichis intended to effectively distribute traffic and environmental loading in such a waythat underlying unbound layers are not exposed to excessive stresses and strains. Thisoften implies comparatively high stiffness of the base course. However, following theprinciple of performance-specific design of each layer, the base course shouldprimarily show adequate fatigue resistance. Consequently, it is often assumed that foran asphalt mixture, there exists incompatibility between stiffness and fatigue-resistance. This assumption is based on the fact that a stiff base course can be obtainedusing either relatively course aggregate or stiff bitumen, which both increase thesensitivity to high strain levels [8.]. In principle, a base course material is affected byall material-related factors including grading, additives as well as air void and bindercontents. In practice, relatively fine grading [9.], low void content [10.] and high

binder contents are strived for. According to APA [5.], there are several possiblestrategies to achieve a base course exhibiting long-life characteristics. One approach toensure good fatigue life, as determined by a given laboratory test and a given


9/24


9

pavement construction, is to restrict the tensile strain at the bottom of the base coursebelow a certain fatigue limit, under which no damage develops. However, theidentification of such a limit is in practice very difficult.

2.1.4 Unbound materials and foundation

Since the formation and sub-soil often constitute relatively weak materials, it is ofutmost importance that the applied loadings are effectively eliminated by the layersabove. Especially, high vertical strains may require relatively thick layers. In this case,unbound road-base or sub-base layers consisting of uncrushed aggregate or crushedaggregate can be suitable. Generally, unbound materials originate from locallyavailable sources, such as native soil, crushed/uncrushed granular materials and re-used (secondary) material. The type and thickness of the unbound materials used forthe unbound road base layers technically depend on the structure to be designed(traffic loading) and the stiffness of the sub-grade. In some countries, climaticconditions may also require relatively thick pavement structures in order to avoidfrost-related heaves. Furthermore, economic factors, such as availability andtransportation costs, also have a significant impact on the amount of unboundmaterials desired in a pavement. In some cases adequate bearing capacity can beachieved using different stabilization techniques. For example, different types ofcement- or lime-stabilization can be used to improve clays.

2.2 DesignToday, pavement design is normally represented by so-called semi-mechanisticmethods, implying that they partly are based on fundamental engineering principles.Semi-mechanistic procedures consist in principle of a structural response model andassociated performance models (see Figure 3). Response models relate traffic loadingto stresses and strains in the pavement structure, while performance models relatecalculated stresses and strains to rate of deterioration. As indicated in Figure 1, theactual design life is an essential feature of the LLP-concept. In most Europeancountries, a design life of 20 years is adopted. Only in the UK is it explicitly statedthat pavements should be designed for 40 years when economically viable [4.].

Even though a number of potential deterioration mechanisms exist, the two maindistresses considered in semi-mechanistic methods are rutting, originating in the sub-grade, and fatigue cracking, initiated in the bottom of the asphalt layer. Structuralrutting originating at the sub-grade is considered by limiting the compressive straininduced at the top of the sub-grade, normally as a function of the expected number of

load applications during the design period and the thickness and stiffness of thepavement layers. These two traditional design criteria were originally based onlaboratory testing and field performance of pavements whose layers were relativelythinner than pavements designed for current relatively high traffic levels.

According to a recent European investigation [11.], the importance of traditionalfatigue cracking and rutting in the sub-grade is nowadays questionable. In thisinvestigation, fatigue cracking and rutting in the sub-grade were only considered to beranked as numbers 6 and 9, respectively. Other deterioration mechanisms such asrutting in the bituminous layers and surface-initiated cracking were considered farmore important: ranked as number 1 and 2, respectively. However, even though

distress mechanisms associated with cold climate such as frost heave, wear due tostudded tyres and low-temperature cracking, achieved low ranking (number 10, 11 and12, respectively), they may be significant in some countries such as Sweden.


10/24


10

In contrast to traditional fatigue cracking and rutting in the sub-grade, the otherdeterioration mechanisms (such as rutting in the bituminous layers, crack initiation inthe surface layer), are difficult to address in (the old) semi-mechanistic designmethods. One way of including these mechanisms is therefore to prescribe functional

demands using different simulation methods (functional laboratory tests).

Figure 3. Schematic illustration of semi-mechanistic design method.

Figure 4. Example of pavement structure and the two traditional design criteria.

In general, deterioration, such as fatigue, is within the semi-mechanistic frameworkgenerally obtained from laboratory testing, which is assumed to represent field

Traffic loading Climate Pavement structure

Material properties

Response model

,

Performance model

Fatigue Deformations ?

Base courses

Binder course

Sur ace course

(Un)bound roadbase

Tensile strains

Subgrade

Compressive strains

Boundmaterials

Unboundmaterials


11/24


11

conditions. Fatigue testing is usually performed by applying a cyclic loading (strain-or stress-controlled), often at a given test temperature and frequency, to an asphaltmixture specimen. The fatigue results obtained at laboratory testing are subsequentlyrelated to the strain response obtained from the structural model. Consequently, thelife of a given pavement structure subjected to repetitive traffic loading can be

calculated when the strains in the structure as well as the relationship between thestrains and deteriorations have been established.

2.2.1 Fatigue crackingFatigue is one of the main distress mechanisms considered in flexible pavement designand is generally considered to be caused by excessive tensile strains due to repetitivetraffic loading. The loading results in growth and coalescence of pre-existing orinitiated microscopic defects, which finally cause the asphalt layer to crack.Eventually, the entire bearing capacity of the pavement structure may be jeopardized.An increased asphalt thickness reduces the traffic-induced strains in the bottom of thebituminous base layer and, consequently, increases the fatigue life.

In principle, the design philosophy of long-life pavements does not differ significantlyfrom traditional semi-mechanistic design methods. However, due to the aim ofrestricting structural deterioration to the pavement surface, conservative design criteriashould be employed. In the UK, the design traffic for 40 years is determined in asimilar manner to that of a 20 year life with enlarged traffic multiplier to account forthe greater growth in traffic over the extended period. Pavements design above atraffic level of 80 million standard axles (80 kN) are said to be designed as long-lifepavements [4.]. The rationale for this traffic level is supported by empirical andanalytical studies [2. and 4.].

In order to eliminate structural deterioration, some investigations have suggested thatconventional transfer functions (i.e. performance models, cf. Figure 3) should bereplaced by explicitly expressed conservative thresholds for allowable strain values. For example, Monismith and Long [12.] suggested that the maximum tensile strain atthe bottom of the asphalt layer should be restricted to 6010 -6. It should beemphasised that the threshold level given by Monismith and Long may not beuniversally applicable since they are derived using a given pavement system, designmethod, bituminous mix and laboratory test equipment. However, in principle itshould be possible to derive corresponding threshold values for other pavementconfigurations and values in the range of 50-7010 -6 m/m have been reported as

fatigue thresholds [13., 14. and 15.]. For the sake of convenience, it should beemphasised that any design parameter and criterion given should not be taken out ofcontext since they are closely linked to a specific design guide and testing methodused.

2.2.2 Permanent deformationThe other major deterioration mechanism considered in semi-mechanistic design isrutting, originating in the sub-grade. As in the case of fatigue cracking (see Section2.2.1), most semi-mechanistic methods are equipped with performance models forcalculating permanent deformations using transfer functions.

In order to eliminate structural deterioration, some investigations have suggested thatconventional transfer functions (i.e. performance models, see Figure 3) should bereplaced by explicitly expressed conservative thresholds for allowable strain values.


12/24


12

For example, Monismith and Long suggested that the maximum vertical compressivestrain at the top of the sub-grade should be restricted to 20010 -6 m/m.

2.2.3 Other mechanismsTraditionally, fatigue cracking has been viewed as initiated in the bottom of the

asphalt layer to propagate up through the layer (bottom-up cracking). However, thereare other forms of fatigue-related cracking than traditional bottom-up cracking, such asreflective cracking . The latter deterioration mechanism is caused by repetitiveshearing, e.g. when a new asphalt layer is laid upon an already cracked layer. Withtime, the crack will propagate through the new layer. A third type of fatigue is surfaceinitiated (top-down) cracking . This type of fatigue cracking is considered to be causedby a complex combination of pavement structure, load spectra and materialcharacteristics. Even though still disputed, this deterioration mechanism motivates acertain fatigue-resistance of the surfacing. In addition to fatigue cracking and rutting,some semi-mechanistic methods also take low-temperature cracking and frost heave into account. Another deterioration mechanism which should be accounted for isageing . This distress mainly affects the bituminous layers and is manifested byincreased stiffness and decreased flexibility. According to Nunn et al [2.] a road ismore susceptible in its early life before the structural properties have been increasedby curing. In this case, curing implies increased bearing capacity due to gradualincrease in asphalt stiffness. However, whether or not increased stiffness is desirabledepends on whether, or not, the stiffening extends pavement life and lowers overalllife-cycle cost. In certain cases, it is not unlikely that ageing will result in lowtemperature cracking but also complicates recycling. Consequently, it will in manycases be desirable to minimize ageing by reducing the air void content of the mixture.

A common denominator of the distress mechanisms mentioned in this section is thatthey are difficult to take into account using semi-analytical methods. Some of thedistresses are considered to require more advanced response and/or performancemodels (see Figure 3). In the case of top-down cracking and permanent deformationsin the bitumen-bound layers, new and improved design methods may take this intoaccount in the future.

However, functional demands may be used to take these mechanisms into account. Forexample, the adequacy of a binder course can be evaluated using a simulation method.If a material shows superior resistance against a certain distress mechanism simulatedby the test it may be judged appropriate to be used in the design.

2.2.4 Improved design methodsToday, there are several design methods available for pavement design.At the National Center for Asphalt Technology (NCAT) at Auburn University inconjunction with the Asphalt Pavement Alliance (APA) a procedure was developedfor the mechanistic-based design of flexible long-life or perpetual pavement structures.This is a program called PerRoad. [19.]

The design software utilizes layered elastic theory to compute critical pavementresponses under axle load spectra. Monte Carlo simulation is used to model theuncertainty corresponding to material, loading and construction variability. The

program can be used as a design and analysis tool to assess the likelihood that criticalpavement responses will exceed a threshold set by the analyst. Additionally, transfer


13/24


13

functions may be used to determine a damage accumulation rate for pavementresponses exceeding the threshold.This software uses U.S. customary units, commonly known in the United States asEnglish units .

2.3 ConstructionLike traditional pavement construction, the construction of LLP should at least beexecuted in accordance with public requirements. In this case, technical and functionalrequirements on LLP should in general be higher compared to traditional pavements.One way of ensuring higher quality is to prescribe functional demands instead oftechnical specifications. This emphasises the contractors roll in achieving long-lifeperformance compared to when technical specifications are used. When functionaldemands are used, it is up to the contractor to show that a solution meets therequirements. At least two benefits can be identified:- improved flexibility, meaningthat several solutions may be considered, and an increased concern from the contractorto optimize the production, e.g. compaction.

The transportation of the asphalt mixture from the asphalt plant to the construction siteis important for the durability of the pavement. During transport the temperature of theasphalt mixture should not drop too much and one should be aware of possiblesegregation. During transportation there could occur temperature segregation and/ormixture segregation. Material Transfer Vehicles that remix the mixture could be usedto avoid temperature and mixture segregation. Adequate quality control duringconstruction is essential too. In this way all the relevant issues will be monitored.

One of the most important construction-related aspects is mixture composition. Anoptimal recipe should result in a homogenous material structure which facilitatesadequate compaction. Results presented in [7, 21] indicate that 3 percent morecompaction results in 15 percent less thickness needed due to improved fatiguebehaviour (in this case 8% air voids versus 5% air voids). Low void and high bindercontents will also improve fatigue characteristics as well as enhance resistance againstmoisture-related damage, which is important since the base course layer often is inprolonged contact with water. Perhaps the most advocated material-relatedimprovement is to modify the bitumen using different polymers. Many studies haveindicated that polymer modified bitumen produces pronounced improved fatiguebehaviour of a given mixture type (e.g. [16.]). The combination of a thick well-constructed pavement and a fatigue-resistant base course is considered sufficient to

guard against traditional fatigue cracking. In the context of long-life pavements, it is still necessary to periodically replace thesurface course in order to fulfil the aim of avoiding structural deterioration. In the UK,two quite different routes have occurred [2.]. In order to improve durability, thickerasphalt layers have been introduced in order to achieve improved compaction. On theother hand, UK has experienced an increased interest in the use of relatively thin hot-mix surface courses (Thin Layers of Asphalt Concrete). Investigations on motorwaysin the UK have suggested that fatigue and structural deformation do not occur inpavements which have been constructed with sufficient strength to resist structuraldeterioration. The main conclusions suggest that well-constructed flexible pavements

show very long service life provided that distress mechanisms such as rutting andcracking are confined to the pavement surface [2. and 3.].Asphalt mixtures areviscoelastic composite materials which show significant time- and temperature-


14/24


14

dependence. At high temperatures and/or low loading rates (vehicle speed), asphaltmixtures are relatively soft, while the opposite applies at low temperatures and/or highloading rates. Another characteristic of asphalt mixtures is the gradual stiffening withtime, ageing. The adequacy of asphalt mixtures in flexible pavement design dependson the quality and composition of the main constituents: bitumen and aggregates.

Binder properties, such as stiffness and ductility, have a well-documented effect onmixture performance since the stiffness of the binder largely influences the stiffness ofthe mixture. The bituminous binder should be chosen taking into consideration theprevailing conditions concerning aggregate type and regional climate since theycollectively influence mixture performance. In some respects, asphalt binder andmixture properties can be improved using certain additives, e.g. polymers. In additionto binder properties, binder content shows a major impact on mechanical properties ofasphalt mixtures. The aggregate should exhibit adequate properties regarding mainlystability and wear resistance.

2.4 Maintenance

In the past, the preoccupation of road engineers has been mostly with construction ofnew roads. However, as road networks in many countries have been completed, theemphasis has gradually turned to maintenance of the existing network [17.]. Incontrast to relatively short-duration construction projects, maintenance is characterisedas an on-going long-term process over a relatively widespread area.

Maintenance of traditional roads normally covers a wide range of activities aimed atreducing the rate of pavement deterioration and in the end achieve low vehicle-operating costs. As more advanced maintenance activities, such as reconstruction, aresignificantly more expensive than resurfacing, it is important to continuously carry outadequate activities in order to avoid an increased deterioration rate. A third reason forcarrying out maintenance is to keep the road open continuously. Any failure to keep ahighly trafficked road open will lead to serious social and economic consequences.For this reason structural damage has to be avoided; repairing structural damage iscostly and time consuming. In contrast to traditional pavements, long-life pavementsare defined as pavements, which do not show any structural deterioration.Consequently, the end of a pavements life will not be reached as long as correctsurface maintenance is carried out.For motorways with a very high traffic intensity the availability of the road is ofutmost importance and therefore disruption to traffic has to be avoided. The

maintenance or the renewing of the surface layer has to be done in a short time periodto minimise the hindrance to the traffic. Besides that the cost of traffic managementmeasures during maintenance works on motorways are high and can be more then50% of the total job costs.The lower need for more extensive structural maintenance activities with LLPcertainly affects the pavements overall life-cycle costs.


15/24


15

3. Economical aspects

As discussed in Section 1, the overall purpose of LLP is to achieve lower annual costsby increased pavement life and fewer and more cost effective maintenance activities.In order to assess the economic benefits of LLP, the full financial consequences

should be considered using Cost/Benefit Analysis (CBA) [4.]:

Initial construction costs The loss of pavement capital due to pavement deterioration Agency costs due to periodic maintenance and traffic management at road

works Road user costs, primarily due to delays at road works The costs due to accidents involving road users and workers at road work

sites The environmental economical impacts of road construction and

maintenance

Intuitively, a LLP probably results in an increase in construction cost compared totraditional pavements, but lower maintenance-related costs. Some of the costsincluded in CBA are relatively simple to determine while others are relativelydifficult. In the case of pavement deterioration, only limited information regardingcosts has been published. The most likely explanation is that present PavementManagement Systems (PMS) mostly do not include residual value of pavements [4.].Costs associated with road user delays can be obtained using models based onmeasured traffic flows and road capacities. Road safety costs are more difficult toappreciate, primarily due to lack of relevant data. The three main areas of concern for

environmental costs are recycling of pavement materials, pollutant impacts related tofuel consumption and noise impacts related to maintenance work. The agency costsconsist mainly of direct costs including maintenance and traffic management costs butalso indirect costs in terms of administration. The review presented by FEHRL [4.]showed that only the UK takes special account of the economic aspects of LLP. UsingCBA, UK experience has shown that pavements classified as LLP are expected to bemore cost-effective than traditional road pavements since the small increase inconstruction cost is compensated by lower direct maintenance costs and indirectdisruption-related costs. For a 10 year period, the total savings by adopting the LLPprincipal in the UK, are expected to reach 350 Million Euro [4.].

However, even though many costs associated with CBA are abstract and difficult toestablish, there are simpler ways of estimating the economic benefits of a LLP.Perhaps the easiest way is to compare the construction and maintenance costs of two(or more) types of pavement structures, one being a LLP and the other being aconventional one. This in turn requires that the design models used are capable oftaking into account the different parameters responsible for the deterioration.

The traditional Life Cycle Cost Analyses (LCCA) can be used for calculating thepresent worth of costs for pavement alternatives and it is the primary tool used foreconomic comparisons. For these LCCA studies the time span that is taken into account is mostly between 20and 40 years. When dealing with Long Life Pavements this time span should be longerand might have to go up to 100 years.


16/24


16

In [20.] Haas et al. suggest a framework for LCCA applications which recognizesshort, medium and long term cycle periods, functional class of highway, public andprivate sectors and likely or preferred LCCA method. Reasonable life cycle periods forshort, medium and long term analysis are in the order of 25, 50 and 100 yearsrespectively. This paper [20.] also gives a numerical example which shows how an

agency could calculate an internal rate of return (IRR) for a base investmentalternative involving a long life pavement design.

4. Future developments

4.1 Procurement and specificationsAn important aspect, which affects the final result in the field, is the means by whichquality is defined, measured, controlled and ensured. In recent years, performancespecifications for pavements have been debated intensively. Performancespecifications are often considered to recognize the relationship between constructionquality and long-term performance. By rationally controlling variables that impactlong-term performance, the quality of the final product can be improved. In order toachieve long-life performance, aspects such as these should be considered in additionto the technical demands.

4.2 Improved economic modelsAs indicated in Section 3, improved economic models are necessary to show thebenefit of alternative designs and materials. Perhaps the most significant obstacle isthe need to develop a model which is capable of illustrating the relation betweeninvestment cost and life-cycle costs. Today, most pavements are purchased primarilybased on investment cost, which consequently, often leads to high life-cycle costssince the procurement procedure rewards the lowest bidding price.

5. Summary and conclusions

During the last couple of years Long-Life Pavements- and Perpetual Pavementconcepts have been established in Europe and in the USA. The main purpose of theseconcepts is to develop a framework, in which more cost-effective pavements can beproduced.When the availability of road lanes and the road user delay costs are taken into accountin the cost-benefit analyses of pavements it will show that a number of highways indensely populated areas require low maintenance pavements. That means a totalreconstruction of a road pavement is hardly possible or even impossible.This is one of the reasons for developing the Long-Life Pavement concept.

A long-life pavement is defined as a type of pavement where no significantdeterioration will develop in the foundations or the road base layers provided thatcorrect surface maintenance is carried out.

This definition implies that all pavement layers, except the road surface layer are

considered as permanent pavement layers.


17/24


17

The LLP-concept is to some extents based on earlier concepts but mainly on recentadvances in material technology, design and functionality.In practice, the intention of the LLP-concept is to significantly extend currentpavement design life by restricting distress, such as cracking and rutting to thepavement surface. Common distress mechanisms such as bottom-up fatigue cracking,

rutting in the unbound layers and frost heave should, in principle, be completelyeliminated.

For design and construction of new pavements this concept means that the pavementhas to be designed and constructed in such a way that

bottom-up fatigue cracking should be prevented rutting in the pavement layers should be avoided by using adequate asphalt

mixtures rutting in the unbound layers and frost heave should completely eliminated.

The choice of the surface layer depends on the functional requirements. This could bea combination of comfort, durability, stability, skid resistance and noise reduction.There might be even additional requirements like surface water drainage or waterimpermeability.A wide range of bituminous surface layer products can be considered appropriatedepending on specific requirements. The selection of surface course is a matter ofidentifying the most appropriate materials during the design life.The bituminous binder courses should be able to withstand high shear stresses and thebituminous base layer should be thick and stiff enough to distribute the traffic loadingsto the substrate and to withstand fatigue.

The resistance to fatigue depends on the fatigue resistance of the bituminous mixtureand the tensile strain at the bottom of the asphalt base course. The tensile strain shouldbe below a certain fatigue limit, under which no damage accumulates.There are two approaches to achieve a sufficient low strain level. On approach is basedon a certain number of Equivalent Standard Axle Loads in the design and the otherapproach is to limit the tensile stain at the bottom of the bituminous base layer to acertain number:

Pavements design above a traffic level of 80 million standard axles (80 kN) aresaid to be designed as long-life pavements.

Threshold values for tensile stain in the range of 50-7010 -6 m/m have beenreported as fatigue thresholds.

In practice both approaches will lead to a tensile stress level at the bottom of thebituminous base layer that is low enough to avoid crack initiation / crack propagation.

It should be emphasised that any design parameter and criterion should not be takenout of its context since they are closely linked to a specific design guide, test methodsused, materials used and the construction of the pavement, etc.In the same way the maximum tensile strain at the top of the sub-grade could bementioned (20010 -6 m/m) to avoid rutting in the top of the subgrade.

In general the technical and functional requirements for LLP will in general be highercompared to traditional pavements. A way to achieve this is by prescribing functionalspecifications instead of technical specifications. Then it is up to the contractor toshow that a solution meets the requirements. In this way several solutions may beconsidered by the contractor and the contractor will optimize the construction.


18/24


18

The traditional Life Cycle Cost Analyses can be used for calculating the present worthof costs for pavement alternatives and it is the primary tool for economic comparisons.For Long-Life Pavement the recommended life cycle periods for short, medium andlong term analysis are in the order of 25, 50 and 100 years respectively.

6. Literature

[1.] Newcomb, D. E., Buncher, M., Huddleston, I. J. Concepts of perpetualpavements. Transportation Research Circular, No 503, Perpetual BituminousPavements, (2001).

[2.] Nunn, M. E., Brown, A., Weston, D., Nicholls, J. C. Design of long-life flexiblepavements for heavy traffic. Highways Agency, TRL report 250, (1997).

[3.] Nunn, M., Long-Life Flexible Roads, TRL, ISAP Conference 1997, Seattle(1997)

[4.] FEHRL A guide to the use of long-life fully-flexible pavements ELLPAGPhase I. FEHRL report 2004/01, (2004).

[5.] APA Perpetual Pavements-A Synthesis. Asphalt Pavement Alliance (APA),published at the Asphalt Pavement Alliance homepage:[email protected], (2002).

[6.] TRB Perpetual Bituminous Pavements. Foreword, Transportation Research

Circular, number 503, December, (2001).

[7.] Harvey, J., Long-Life AC Pavements: A Discussion of Design and constructionCriteria Based on California Experience Paper 12 Proceedings InternationalSymposium on Design and Construction of Long Lasting Asphalt Pavements,Auburn USA, 7-9 June 2004.

[8.] Tseng, K. H., Lytton, R. L. Fatigue damage properties of asphaltic concretepavements. Transporta-tion Research Record 1286 (TRB), pp 150-163, (1990).

[9.] Sousa, J. B., Pais, J. B., Prates, M., Barros, R., Langlois, P., Leclerc, A. M.Effect of aggregate grada-tion on fatigue life on asphalt concrete mixes.Transportation Research Record 1630, (TRB), pp 62-68, (1997).

[10.] Tayebali, A. A., Deacon, J. A., Coplantz, J. S., Finn, F. N., Monismith, C. L.Fatigue Response of Asphalt, Part II-Extended Test Program.SHRP Project A-404, Asphalt Research Program, Institute of Transportation Studies, Universityof California, Berkeley, (1994).

[11.] Development of New Bituminous Pavement Design Method, Final report ofCOST Action 333, European Commission, Luxembourg, 1999.

[12.] Monismith, C. L., Long, F. Overlay Design for Cracked and Seated PortlandCement Concrete (PCC) Pavement-Interstate Route 710. Technical


19/24


19

Memorandum TM UCB PRC 99-3, Pavement Research Center, Institute forTransportation Studies, University of California, Berkeley, (1999)

[13.] Timm, D.H. The Effects of Load Spectra and Variability of Perpetual PavementDesign Paper 7 Proceedings International Symposium on Design and

Construction of Long Lasting Asphalt Pavements, Auburn USA, 7-9 June 2004.

[14.] Thompson, M., Design Principles for Long Lasting HMA Pavements Paper 14Proceedings International Symposium on Design and Construction of LongLasting Asphalt Pavements, Auburn USA, 7-9 June 2004.

[15.] Peterson, R., Determination of Threshold Strain Level for Fatigue EnduranceLimit in Asphalt Mixtures Paper 15 Proceedings International Symposium onDesign and Construction of Long Lasting Asphalt Pavements, Auburn USA, 7-9June 2004.

[16.] Lundstrom, R. Isacsson, U. An Investigation of the Applicability of SchaperysWork Potential Model for Characterization of Asphalt Fatigue Behavior.Published and Presented at the annual AAPT meeting, Baton Rouge, USA, mars2004. (2004)

[17.] Robinson, R., Danielson, U., Snaith, M. Road Maintenance Management Concepts and Systems. Macmillan Press LTD, (1998).

[18.] Lundstrom, R., Ekblad, J., Isacsson, U., Karlsson, R. Fatigue Modeling asrelated to Flexible Pavement Design-State of the Art. Accepted for publicationin International Journal of Road Materials and Pavement Design, (2006)

[19.] Newcomb, D.E., A Guide to PerRoad 2.4 for PerRoad Perpetual Softwaredeveloped by Professor David H. Timm. Asphalt Pavement Alliance, August2004. http://www.asphaltalliance.com/

[20.] Haas, R., Tighe, S. and Cowe Falls, L., Return on Investment Analysis for LongLife Asphalt Pavements Proceedings 11 th International Conference on asphaltPavements, August 12 17, 2006, Qubec City, Canada.

[21.] Harvey, J., Long, F., and Prozzi, J.A. Application of CAL/APT results to LongLife Flexible pavement reconstruction, Accelerated Pavement TestingConference, Reno, NV, USA, 1999


20/24


20

Appendix 1

Durability of surface layers in Europe

A.1. Introduction

Long-life Pavements or Perpetual Pavements mean building roads in such a way thatfatigue at the bottom of the asphalt pavement layer is not a critical parameter. This alsomeans that the total thickness of the pavement layers should be enough to avoid fatiguestarting from the bottom.Consequently, for maintaining the road in a good condition it is sufficient to replace thesurface-layer only. In this way the durability of the surface layer is now the dominantfactor for the pavement surface life.

To get an insight into the service life of surface layers the EAPA Technical Committeeinvestigated the durability of the surface layers used in Europe.

This annex gives an overview of the different asphalt mix types that can be used forsurface layers and their expected durability. The data are based on a questionnaire and apanel discussion of experts.

A.2. Asphalt mixtures

In the descriptions of the asphalt mixtures (1 to 7) the definitions are from the EuropeanAsphalt Standards EN 13108 1 to 7 and they are quoted with inverted commas.

1. Asphaltic Concrete (AC)

Asphalt in which the aggregate particles are continuously graded or gap-graded toform an interlocking structure. Dense asphalt concrete is often used as the basicsurface layer.

2. Asphalt Concrete for very thin layers (AC-TL)Asphalt for surface courses with a thickness of 20 mm to 30 mm, in which theaggregate particles are generally gap-graded to form a stone to stone contact and toprovide an open surface texture. This mixture is often used in France and is calledBBTM (Bton Bitumineuse Trs Mince).

3. Soft Asphalt (SA)Mixture of aggregate and soft bitumen grades. This flexible mixture is used in theNordic countries for secondary roads.

4. Hot Rolled Asphalt (HRA)Dense, gap graded bituminous mixture in which the mortar of fine aggregate, fillerand high viscosity binder are major contributors to the performance of the laidmaterial. Coated chippings (nominally single size aggregate particles with a highresistance to polishing, which are lightly coated with high viscosity binder) are alwaysrolled into and form part of a Hot Rolled Asphalt surface course. This durable surfacelayer was often used as a surface layer in the UK.


21/24


21

5. Stone Mastic Asphalt (SMA)

Gap-graded asphalt mixture with bitumen as a binder, composed of a coarse crushedaggregate skeleton bound with a mastic mortar. This mixture is of often used as asurface layer in case high stability is needed. The surface structure also has good noise

reducing properties.

6 Mastic Asphalt (MA)Voidless asphalt mixtures with bitumen as a binder in which the volume of filler andbinder exceeds the volume of the remaining voids in the mixed. This mixture is verydurable and was often used as surface layer in certain countries.

7. Porous Asphalt (PA)Bituminous material with bitumen as a binder prepared so as to have a very highcontent of interconnected voids which allow passage of water and air in order toprovide the compacted mixture with drain and noise reducing characteristics.

8. Double layered Porous Asphalt (2L-PA)The top layer with a grain size 4/8 mm is about 25 mm thick and the second/bottomlayer is porous asphalt with a course aggregate (11/16 mm). The total thickness is about70 mm. Because of the finer texture at the top (that gives less tyre vibrations), it gives abetter noise reduction than a single layer porous asphalt.

9. Asphalt Concrete for ultra thin layers (UTLAC)Asphalt for surface courses with a thickness of 10 mm to 20 mm, in which theaggregate particles are generally gap-graded to form a stone to stone contact and toprovide an open surface texture. Several varieties of this layer are often used to providea good, new noise reducing surface layer.

The difference between the mixtures with regard to structure and skeleton are alsoillustrated in figures 5, 6 and 7.

Figure 5: Difference in structure between AC, SMA, BBTM and PA.


22/24


22

Figure 6: Difference between sand skeleton and stone skeleton mixtures

Figure 7: Structure of double- layered Porous Asphalt

A.3. Basic assumptions for determining the durability of the surface layers

The expected durability is based on the best practice. This means that surface layers areexpected to be laid on a properly designed road where the durability of the surface layer isnot determined by bottom-up fatigue cracking.Also, it is assumed that the mix is properly designed and well compacted. As an example,a SMA surface layer that shows rutting in the SMA is an improperly designed mixture.

The data shown are not valid in the case of using studded tyres.The durability of surface layers also depends on the local conditions, the local climate, themix formula, the types of bitumen and aggregate used, the maximum allowable axle loadsand the actual axle loads used (because excessively high axle loads have a strong negativeinfluence on the service life of the pavement surface layers). In some cases or countriesthe skid resistance of the surface layer is the dominant factor for determining the servicelife, so the durability depends on the required friction level and the PSV-value ofaggregate that has been used.

A distinction has been made between major roads (motorways and heavily traffic roads)and secondary roads.


23/24


23

A.4. Results

Table 1 and Figure 8 show the Durability of Surface Layers is expressed in years ofservice life for major roads / motorways / heavily trafficked roads.

Table 2 and Figure 9 show the Durability of Surface Layers is expressed in years ofservice life for secondary roads.

Major roads / motorways / heavily trafficked

Type 15% Lowel level European average 85% Higher levelAC 8 14 18AC-TL 30-40mm 8 12 18AC-VTL 25-30mm 8 10 12UTLAC 8 10 12

PA 8 10 142L-PA 1) 9 11 12SMA 14 20 25HRA 17 21 25Mastic-A 18 21 24

Soft-A is not used here1) Only based on Dutch results

Table 1: Durability of surface layers on major roads

Major roads / Motorways / Heavily trafficked roads

0

5

10

15

20

25

30

AC AC-TL AC-VTL UTLAC PA 2L-PA SMA HRA Mastic-A

Surface layer type

Y e a r s

15% Lowel level

85% Higher level

European average

Figure 8: Durability of surface layers on major roads


24/24


Secondary roadsType 15% Lowel level European average 85% Higher levelAC 10 15 20

AC-TL 10 15 20AC-VTL 10 12 142L-PA 1) 10 11 12SMA 16 20 25HRA 20 25 30Mastic-A 18 24 30Soft-A 8 12 25

PA and UTLAC are not used here1) Only based on Dutch results

Table 2: Durability of surface layers on secondary roads

Secondary roads

0

5

10

15

20

25

30

35

AC AC-TL AC-VTL 2L-PA SMA HRA Mastic-A Soft-A

Surface layer type

Y e a r s 15% Lowel level

85% Higher levelEuropean average

Figure 9: Durability of surface layers on secondary roads

Documents

Long-life Asphalt Pavements_technical2007